Page:Transactions of the Geological Society, 1st series, vol. 2.djvu/280

270 easily conceived to have arisen from a partial neglect of the fire after the wall had nearly attained its requisite height; nor is there any reason why it should not have been increased in height by the addition of cold stones after a firm foundation had been obtained.

One other circumstance in the appearance of the burnt stones is deserving of notice before quitting this subject. The changes which the mica slate has undergone, appear to be such as could not have been produced but by long torrefaction, or by such a repetition of the heat as I have supposed to be the result of design. The transient effects which would follow from burning down a wooden wall, would scarcely have been sensible on stones of so refractory a nature, which exhibit changes in many instances as great as if they had been exposed for a long time to the heat of an ardent furnace.

Such are the observations to which a consideration of the fort of Dun Mac Sniochain has given rise.

As this was the only one of these mysterious fabrics which I had seen when the above remarks suggested themselves to me, I was afterwards glad to have an opportunity of examining the fort on Craig Phadric, it being that one on which most labour had been bestowed, and that which I thought might possibly either confirm or refute my notions on the subject.

Its general appearance and military structure having been fully and carefully described, I shall only indulge in a very few remarks on its physical composition.

The hill of Craig Phadric, on which it stands, is one of a numerous set of pudding-stone rocks, which may be traced from Fyers, and for aught I know, beyond it. At Fyers they lie above the primary rocks, which they doubtless separate as usual from the secondary strata, as they may be seen near Inverness, succeeded by sandstone breccia and common sandstone. On the top of the rock there is a deep deposit of rounded stones, consisting of fragments of the older